People wait in line to meet with recruiters during a job fair in Philadelphia. (AP/Matt Rourke)

Just looking at the economy's overall size, you wouldn't think that the last year was much different from any of the others since the recession. The U.S. economy grew at about the same rate in 2014 as it did in the previous four years -- less than 2.4 percent, according to the Federal Reserve's most recent projection. Yet last year was different. People started going back to work. The percentage of Americans working, more or less stuck in a ditch since 2009, increased from 58.6 percent in December 2013 to 59.2 percent last month. Employers added an average of 246,000 positions a month, about 3 million jobs overall.

Economists will debate what happened, but one of the more controversial theories is that Congress's decision not to extend federal unemployment benefits at the end of 2013 encouraged those out of work to settle for more poorly paid jobs, giving firms a better reason to expand and hire new workers. That's the conclusion of a new working paper from the National Bureau of Economic Research. The authors, Marcus Hagedorn of the University of Oslo, Iourii Manovskii of the University of Pennsylvania and Stockholm University's Kurt Mitman concluded that the reduction in benefits created 1.8 million jobs last year -- more than half of the total. Their work has already drawn several rebuttals from other economists criticizing their choice of methods and data.

Congress first extended unemployment benefits during the financial crisis as a way of helping people who had lost their jobs get by. The program was implemented in different ways, though, so that by the end of 2013, someone who lost her job in North Carolina could rely on only 19 weeks of unemployment benefits, while someone in Nevada would get 73 weeks. Republican lawmakers refused to extend the program again last year, despite the warnings of many economists, including President Obama's advisers, that failing to do so would hamper the economy. They argued that the unemployed spend the money they receive in benefits, creating demand for goods and services.

Hagedorn, Manovskii and Mitman argue that the opposite occurred. Not only did the economy put more people back to work overall, but the increase in employment was especially pronounced in states like Nevada that had generous unemployment benefits before. Looking at comparisons between neighboring counties on the opposite sides of state lines, where you'd expect the economy to be more or less the same, they found a similar result: The greater the reduction in benefits, the greater the increase in employment.

The group also estimated that about half of those new jobs were filled by people who were previously not in the labor force -- students or people who had given up on their careers and then decided to come out of retirement.

That said, the results are not a vindication for Republicans like Sen. Richard Shelby (R-Ala.). "People, if you pay 'em for years and years, they won't look for a job," he said, reiterating a common conservative argument against unemployment insurance.

In an interview, Mitman rejected "the idea that people just collect the check and sit at home and watch T.V." People who are out of work are always out looking for jobs, whether or not they're getting unemployment insurance, he said.

Instead, because they were receiving checks, the unemployed could afford to wait until they found a job with a higher wage. It wasn't workers who gave up, the group argues -- it was employers, who found that they had to pay more to bring in new labor. After a while, they were less likely to advertise new vacancies.

In a previous study, the authors did find a correlation between higher wages and unemployment benefits, which suggests that those out of work were able to demand more from employers when they had insurance.

Other economists remain skeptical of the group's work. "I have real concerns about the last study. I continue to have those concerns. I don't think they've been resolved," said Jesse Rothstein of the University of California, Berkeley.

He pointed out that the researchers relied on data for individual counties that was partly estimated based on data from the entire surrounding state. Neighboring counties where conditions for workers were in fact quite similar may have appeared to be very different based on these estimates, because they were in states with divergent economic fortunes.

Since the federal insurance program provided more generous unemployment benefits in states with higher rates of unemployment, Congress reduced benefits most drastically in the states with the worst economies. Data from counties in those states would reflect the poor economic condition of the state as a whole. And since economies tend to bounce back harder the farther they've fallen, the fact that employment improved more quickly in those counties with drastic cuts in unemployment might simply result from the fact that the statewide situation was worse initially.

In response to the paper, Dean Baker, a director of the liberal Center for Economic and Policy Research, conducted a similar analysis using a different set of data and came to the opposite conclusion as Hagedorn, Manovskii and Mitman.

Mitman defended the group's methods. He said the combination of sources in the data they used helped reduce noise and better captured the formation of new businesses.

Even if their conclusions are correct, it's worth noting that unemployment insurance has likely still helped those who were laid off. Without benefits, the paper finds, they were forced to accept lower wages for their labor.